HyperScience: Harnessing the Power of AI to Transform Insurance Data and Processes
HyperScience uses the latest in Machine Learning (ML) to automatically classify and extract data from the billions of documents – application forms, enrollment documents, claims, etc. – that flow between insurance companies and their customers and partners every year. “We help the world’s leading organizations harness the power of automation to unlock data, achieve efficiencies, improve response times, and elevate customer experience,” says Peter Brodsky, the chief executive officer and co-founder of HyperScience. “Organizations have access to more data with fewer errors and at lower costs, employees can focus on activities that drive the business forward, and most importantly, customers get the service and answers they expect and deserve.” By leveraging the latest AI tech in HyperScience, insurance companies can automate mission-critical processes and ensure the accurate, complete data they need to assess risks better, underwrite policies, process claims, and deliver timely (often critical) services to their customers.
Despite being “digital-first,” the mission-critical task of classifying and processing documents is still a very manual, expensive process for organizations today. Businesses are spending as much as $60 billion each year on data entry, and that figure is only getting larger. HyperScience’s proprietary Machine Learning models automatically classify and extract data across diverse inputs (such as handwritten forms or low resolution, distorted images) with greater accuracy than any other solution in the market today. Structured data files can be sent to downstream systems for quicker processing, resulting in improved customer experience and faster time-to-revenue. In addition to being able to tackle diverse document types, three key factors make the company’s system easier to use than other solutions: 1) how it measures confidence, 2) how it incorporates humans-in-the-loop, and 3) its user interface, which is designed for non-technical business users.
HyperScience’s built-in quality assurance mechanism is exceptionally good at knowing when it is likely to be right as well as when it’s likely to be wrong. When it is less sure of a transcription, it sends the field to human supervisors to review and resolve. This fine-tunes the underlying models. In addition, the entire system is built with the need for human feedback in mind. “When our system is not sure how to proceed, we involve humans to help, sending a subset of exception cases to data keyers to review and resolve. This provides statistically significant data on both machine and human transcription and improves overall model performance,” mentions Brodsky.
We deliver an automation solution that enables more accurate d ata, increased capacity, faster response times, and improved customer experience
Since closing its Series B in January 2019, HyperScience has surpassed the 115 employee mark, opened its second European office in London to fuel international expansion, and consistently achieved double-digit growth month-over-month. HyperScience is the platform of choice for leading insurance, financial services, healthcare and government organizations worldwide, including TD Ameritrade, QBE, and Voya Financial. On the product side, in 2019, HyperScience took significant steps towards its ultimate vision of making the platform input-agnostic, capable of extracting data from every document type (i.e., any structure and language) and flexible enough to adapt to any processing workflow. In addition to adding language support for French, Spanish, and German, the company continues to improve their specialized models to maximize extraction automation and accuracy for semi-structured documents.
For the future, HyperScience will continue to focus on serving customer needs and expanding its presence internationally. “We make it a priority to invest in our product and engineering, and will continue to do so, so that our research & development team has the resources they need to stay at the forefront of an ever-changing field,” concludes Brodsky.